Keras_and_IMDB-2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Keras and IMDB.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "[View in Colaboratory](https://colab.research.google.com/github/bhavya5491/Machine-Learning/blob/master/Keras_and_IMDB-2.ipynb)"
      ]
    },
    {
      "metadata": {
        "id": "Q4W7ZU4uIG6v",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from keras.datasets import imdb"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Lb5h4XUtIJfz",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "(train_data, train_labels), (test_data, test_labels) = imdb.load_data(num_words=10000)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "HCCs65PgIeya",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "06791f86-695c-4560-c2f7-0bf6c55390aa"
      },
      "cell_type": "code",
      "source": [
        "\n",
        "len(test_data[10])"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "209"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "metadata": {
        "id": "Mab3IylvIlJI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "9502020e-5a90-4439-a25e-7e65e832c2e1"
      },
      "cell_type": "code",
      "source": [
        "len(train_data[10])"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "450"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "metadata": {
        "id": "YqZ0JWVSIvqn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bf473e03-b05d-4f7b-b9ad-b5aeb8a3a154"
      },
      "cell_type": "code",
      "source": [
        "train_labels"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 0, 0, ..., 0, 1, 0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 32
        }
      ]
    },
    {
      "metadata": {
        "id": "UOxSd_awJU_Y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f98f89bd-bec5-425e-9af9-12dc6d7a06a1"
      },
      "cell_type": "code",
      "source": [
        "test_labels"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 1, 1, ..., 0, 0, 0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        }
      ]
    },
    {
      "metadata": {
        "id": "kxITPtaZJbdB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "75e565d2-3ada-4ce6-fdc8-fbcd61edbcef"
      },
      "cell_type": "code",
      "source": [
        "max([max(sequence) for sequence in train_data])"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9999"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        }
      ]
    },
    {
      "metadata": {
        "id": "3uSnSWSqKL8G",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3ed0d38b-7771-4dce-bf99-15bad24e7c03"
      },
      "cell_type": "code",
      "source": [
        "max([max(sequence) for sequence in test_data])"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9999"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 35
        }
      ]
    },
    {
      "metadata": {
        "id": "5abN4s68KaxG",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import numpy as np"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "BvEdEk5UKrDL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "80be4b94-bf80-4ace-f56e-c2d833e7b988"
      },
      "cell_type": "code",
      "source": [
        "# Using one hot encoding manually creating 10,000 dimensions\n",
        "def vectorize_sequence(data, dimension=10000):\n",
        "  results = np.zeros((len(data), dimension))\n",
        "  for i, sequence in enumerate(data):\n",
        "    results[i, sequence] = 1\n",
        "  return (results)\n",
        "x_train = vectorize_sequence(train_data)\n",
        "x_test = vectorize_sequence(test_data)\n",
        "\n",
        "x_train.shape"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 10000)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "metadata": {
        "id": "mlljGddSLeRX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "114b4917-1f8f-471b-90ae-2c1d5ad7f390"
      },
      "cell_type": "code",
      "source": [
        "x_test[0]"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0., 1., 1., ..., 0., 0., 0.])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "metadata": {
        "id": "adH67nIeL6rN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f3fb75b7-9e8e-4bad-d410-d7fc9a0e5631"
      },
      "cell_type": "code",
      "source": [
        "y_train = np.asarray(train_labels).astype('float32')\n",
        "y_test = np.asarray(test_labels).astype('float32')\n",
        "y_test.shape"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        }
      ]
    },
    {
      "metadata": {
        "id": "kIMm5tojMWkj",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Initializing model and layers\n",
        "from keras import models\n",
        "from keras import layers"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7Dj7Ao2wetw_",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model = models.Sequential()\n",
        "model.add(layers.Dense(16, activation='relu', input_shape=(10000,)))\n",
        "model.add(layers.Dense(16, activation='relu'))\n",
        "model.add(layers.Dense(1, activation='sigmoid'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fax52YlWflpu",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#Compiling the model which inlcudes optimizer, loss and metrics used\n",
        "model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Znst_OuPf3eV",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#from keras import optimizers\n",
        "#model.compile(optimizer=optimizers.RMSprop(lr=0.001), loss='binary_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rwt8zbHtgUTN",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#from keras import losses\n",
        "#from keras import metrics"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "O7sHYE_YglpI",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#model.compile(optimizer=optimizers.RMSprop(lr=0.001), loss=losses.binary_crossentropy, metrics=[metrics.binary_accuracy])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "VXbCG4yjg8HN",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Dividing the dataset into train and validation set\n",
        "x_val = x_train[:1000]\n",
        "partial_x_train = x_train[1000:]\n",
        "y_val = y_train[:1000]\n",
        "partial_y_train = y_train[1000:]\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "98fsx6HZhYgV",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "#history = model.fit(partial_x_train, partial_y_train, epochs=20, batch_size=512, validation_data=(x_val,y_val))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "SgIjE3QEiC-f",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#history_dict = history.history\n",
        "#history_dict.keys()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "k7Ctb22FifFF",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#history_dict.values()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xwCHgEq-ihkd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "h8ACe5i3iupf",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#loss_values = history_dict['loss']\n",
        "#val_loss_values = history_dict['val_loss']\n",
        "#acc = history_dict['acc']\n",
        "#epochs = range(1, len(acc) + 1)\n",
        "\n",
        "#plt.plot(epochs, loss_values, 'bo', label='Training Loss')\n",
        "#plt.plot(epochs, val_loss_values, 'b', label='Validation Loss')\n",
        "#plt.title('Training and Validation loss')\n",
        "#plt.xlabel('Epochs')\n",
        "#plt.ylabel('Loss')\n",
        "#plt.legend()\n",
        "#plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ppWsqD33jjI6",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Plotting the training and validation accuracy\n",
        "\n",
        "#plt.clf()\n",
        "#val_acc = history_dict['val_acc']\n",
        "#plt.plot(epochs, acc, 'bo', label='Training accuracy')\n",
        "#plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "#plt.title(\"Training and Validation accuracy\")\n",
        "#plt.xlabel('Epochs')\n",
        "#plt.ylabel('Accuracy')\n",
        "#plt.legend()\n",
        "#plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "yICJGZuGkrsc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        },
        "outputId": "9a38a011-da48-4323-ba64-93e3fd38c262"
      },
      "cell_type": "code",
      "source": [
        "# Case of overfitting as validation accuracy and loss go down after 4 epochs and training accuracy goes to almost 1\n",
        "# Let's buld a new model with 4 epochs\n",
        "model.fit(x_train, y_train, epochs=4, batch_size=512)\n",
        "results = model.evaluate(x_test, y_test)"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/4\n",
            "25000/25000 [==============================] - 2s 89us/step - loss: 0.4479 - acc: 0.8170\n",
            "Epoch 2/4\n",
            "25000/25000 [==============================] - 2s 86us/step - loss: 0.2548 - acc: 0.9098\n",
            "Epoch 3/4\n",
            "25000/25000 [==============================] - 2s 82us/step - loss: 0.1970 - acc: 0.9296\n",
            "Epoch 4/4\n",
            "25000/25000 [==============================] - 2s 82us/step - loss: 0.1673 - acc: 0.9408\n",
            "25000/25000 [==============================] - 2s 76us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "tYU6XoS9lz98",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "30ff3c92-9c6a-4493-b17e-ab3515509545"
      },
      "cell_type": "code",
      "source": [
        "results"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.2944350778102875, 0.88368]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "metadata": {
        "id": "ycOM0eK6mPfb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "617a1231-6894-4e40-fa42-9769bb5f4f1b"
      },
      "cell_type": "code",
      "source": [
        "model.predict(x_test)"
      ],
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.2243466 ],\n",
              "       [0.9997379 ],\n",
              "       [0.83296376],\n",
              "       ...,\n",
              "       [0.1945436 ],\n",
              "       [0.0645091 ],\n",
              "       [0.5846836 ]], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        }
      ]
    },
    {
      "metadata": {
        "id": "0oQG_QQUmh7L",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}